Rational development of new anticancer drugs involves the identification of potential targets in tumor cells and the development of reagents selectively acting at such targets. It is now possible to design target- specific "genetic drugs", such as antisense oligonucleotides, peptides, or vectors for gene therapy, directly on the basis of the structure of the selected target gene. Some prototype genetic drugs targeting specific oncogenes were reported to have a selective cytostatic or cytotoxic effect on neoplastically transformed relative to untransformed cells. Tumor- specific changes in the structure or expression of known oncogenes, however, do not provide all the determinants of tumor specificity for either conventional or genetic anticancer drugs. The present application proposes to undertake a general search for genetic drugs that would induce selective cytotoxicity or growth arrest in tumor cells, and to identify the cellular targets of such drugs. This search will be based on the new methodology for expression selection of genetic suppressor elements (GSEs), short cDNA fragments that encode peptides or antisense RNA sequences interfering with the function of genes from which they are derived. This methodology was developed in a bacterial model and then used in mammalian cells to isolate a series of GSEs inhibiting human topoisomerase II. In subsequent studies, GSE selection was conducted using a retroviral expression library carrying normalized (uniform abundance) fragments of total cellular cDNA from NIH 3T3 cells. These studies resulted in the isolation of several GSEs derived from previously unknown genes associated with drug resistance or neoplastic transformation. The principle of expression selection of GSEs from retroviral libraries of normalized cellular cDNA will now be combined with the use of a tightly regulated inducible promoter and negative selection techniques, to isolate GSEs of unknown nature, which are cytostatic or cytotoxic to cells in which they are expressed. GSEs that show selective inhibition of transformed relative to untransformed mouse fibroblasts, or that are cytotoxic to human HeLa carcinoma cells, will be isolated and sequenced. The activity of the isolated GSEs will be tested on cell lines derived from different types of mouse or human tumors, and on normal cells. Full- length human cDNA sequences of the genes targeted by the selected GSEs will be cloned and their expression patterns will be characterized. The cloned genes will be used individually to isolate the most efficient GSEs interfering with their functions. These GSEs will then be tested for in vitro anti tumor activity, in the form of retroviral vectors or chemically synthesized peptides or oligonucleotides. It is hoped that the proposed studies will lead to the identification of new targets and prototype genetic drugs with potential utility for the treatment of cancer.